The Deadly Dinner Party: and Other Medical Detective Stories

by Dr. Jonathan A. Edlow M. D.

  Control of typhoid fever, however, as for so many infectious diseases, represents a triumph not of antibiotics but of modern sanitation and public health strategies. After the experience in the Spanish-American War, Dr. Walter Reed drafted a report on typhoid fever in which he concluded: “Indeed, the history of this disease justify us in stating that whenever men congregate and live without adequate provision for disposing of their excrement, there and then typhoid fever will appear.” The widespread sanitary practices that were instituted, partly as a result of Reed’s report, led to a marked reduction in the importance of typhoid fever as a serious health threat to soldiers in World War I and subsequent wars.

  Typhoid fever is a classic food- and water-borne disease, spread by what doctors refer to as the “fecal-oral” route. This somewhat euphemistic expression cannot hide the simple fact that new cases of typhoid fever occur as a result of the organism spreading from the rear end of the intestinal tract of one human being to the front end of another, however circuitously. Epidemics have been linked to sewage, but also to milk, ice cream, meats, canned foods, salads, and even shellfish. The typhoid bacillus will grow temporarily in all of these environments, but the bacteria’s only natural reservoir is in people.

  Consequently, every typhoid epidemic can be traced ultimately to a human being—either someone suffering from the disease, or an asymptomatic carrier of the bacillus.

  One notorious and illustrative example of this principle was an outbreak in Aberdeen, Scotland, in 1964. On Tuesday evening, May 19, 1964, two students with gastrointestinal complaints were transferred from the Aberdeen Royal Infirmary to the city hospital with a clinical suspicion of typhoid fever. A woman had been admitted to the same hospital three days earlier with similar symptoms. The day the two students were transferred, that woman’s husband, her two children, and another unrelated boy were also admitted with similar symptoms. The next day, May 20, the first cultures began to return positive for S. typhi, and over the following days dozens more patients arrived with the same problem. After taking a history from each of them, the Aberdeen public health physicians found that eleven of the first twelve patients had eaten canned corned beef that had been purchased at a modern supermarket in the city’s west end.

  Ultimately, 515 patients were afflicted in this epidemic, and 487 were hospitalized, making it one of the largest and certainly one of the most notorious typhoid outbreaks in the twentieth century. Although treated patients often felt improved after days, many were hospitalized for weeks and some for as long as three months in an effort to prevent secondary spread to others. Frequent press conferences were held; the amount of hysteria this outbreak generated was enormous. Aberdeen lost millions of pounds in revenue due to the publicity. Even Queen Elizabeth paid a formal visit on June 28, partly to show her subjects that being in Aberdeen was safe.

  Health inspectors traced the path of the canned corned beef, certain that the ultimate source of infection had to have started with a human being. Knowing the incubation period of typhoid fever, the inspectors knew that the contaminated can must have been opened on May 6 or 7. Canned products are heated to temperatures that sterilize their contents, but after this heating process the cans need to be cooled down. The researchers found that, although we think of cans as completely impermeable, in fact they are not. The rapid swings of temperature lead to pressure changes that can create microscopic leaks in their seams. Initially, when the hot cans are placed in cool water, there is a pressure gradient pushing outward from the cans, but as the contents cool, this gradient can reverse, drawing water into the can. If the cooling water is pure, there is no problem; however, if it is not, impurities can enter the can.

  The Aberdeen corned beef had been imported from Argentina. The canning plant was located on the Parana River, just downstream from the city of Rosario, with a population of more than 600,000 people. The city’s sewage was discharged into the river, and the nearest sewage discharge pipe was about half a mile upstream from the factory. Investigators estimated that every day 66 tons of excrement and 250,000 gallons of urine passed by the plant. The fast-flowing Parana River at Rosario is about 50 feet deep and 1,300 yards wide, so the sewage was considerably diluted. But the incidence of typhoid in Argentina in the 1960s was approximately 500 cases per million inhabitants (in Britain at the same time, it was 2). A single typhoid carrier can excrete millions of bacteria per day. So it is very possible that typhoid bacilli, which are known to survive in river water, were present as the water flowed past the factory. And here, the Rosario canning plant draws its water to cool the cans after they are heat sterilized.

  Incredibly, in 1955, another outbreak of typhoid in Great Britain, at Pickering, was linked to tongue meat that had been canned at the very same plant in Rosario. As a direct result of that epidemic, the Argentine plant began to chlorinate the cooling water. In 1962, however, there had been a breakdown in the chlorination process that, for various economic and logistical reasons, had never been fixed. So despite the history of the Pickering outbreak, at the time of the Aberdeen epidemic, the cooling water from the river was untreated and almost certainly contained S. typhi.

  But if S. typhi was getting into the cans, one would expect them to also contain other bacteria that are far more plentiful, especially Escherichia coli. It is for this reason that another quality control is implemented in the canning process. After the cans are cooled and washed, they are observed for fifteen days. The purpose of this step is to watch for “blown cans.” Any E. coli that has entered the cans will rapidly multiply. This produces gas that pushes out the walls of the cans, making them appear “blown” from the inside. Also, once a can like this is opened, it is usually obvious that the meat has gone bad, so the consumer will smell and see that the food is contaminated, and not eat it.

  This notion was dispelled following the Aberdeen epidemic of 1964, however, when on June 1, some normal-appearing cans of corned beef from Argentina were cultured. Although the meat appeared wholesome and the cans were not blown, both E. coli and S. typhi were found in the meat. The other interesting finding was that only a portion of the meat grew the bacteria, and other samples from the same can were sterile, so it was possible that a small amount of contamination had occurred but the conditions in the can were not conducive to further growth (which would presumably explain why the cans were not blown). Could this have caused the illnesses in Aberdeen?

  An investigation later revealed that, once the original six-pound can of corned beef had been opened at the supermarket, some of the contents were placed in the cold-meats counter, and some in a south-facing display window in the store. Meteorological records showed that it was warm in Aberdeen for the days of May 7– 9, and the meat in the southfacing window provided ideal conditions for S. typhi to multiply. And S. typhi loves to multiply; several bacteria can grow into several million in a matter of hours. The connection between a human source from Rosario, Argentina, and the contaminated food source in Aberdeen, Scotland, had been established.

  Dr. Brodey’s first thought was the lab result on Roy Harvey’s blood had to be a mistake. After all, he had been a practicing infectious diseases specialist for a decade and had never encountered a case of typhoid fever. But then the blood culture from the second patient, whose story sounded so much like Harvey’s, turned positive for S. typhi as well. “At that point,” recalls Dr. Brodey, “I knew something was going on. I reinterviewed both men at length; neither had been out of the country. These second interviews revealed one important connection, however. Both men were firefighters, and both had been to the convention in the Catskills.” Since his experience with typhoid had just rocketed from zero in ten years to two cases in a day, Brodey feared this might be the tip of an iceberg.

  He promptly notified the New York State Health Department in Albany.

  “Like every outbreak,” remembers Stan Kondracki somewhat facetiously, “it was reported on a Friday night.” In 1989, Kondracki was the co-coordinator of the New York state regio
nal epidemiology program. “We had the county people ship the diagnostic specimens from the two patients to the state lab in Albany for confirmation of typhoid. Over the weekend, we got that confirmation and the investigation started Monday morning. It was an epidemiologist’s delight, but required lots of work. The Fire Chiefs’ Association sent us a computer printout of the names and phone numbers of all the conventioneers. The hotels gave us lists of the other guests. There were senior citizens groups from Ohio, Connecticut, and Pennsylvania, as well as individual guests. We drafted staff from the AIDS and sexually transmitted diseases sections to help with all of the interviews and telephone calls.”

  In fact, the team included a large number of individuals not only from the New York State Department of Health’s Bureau of Communicable Diseases but also from its sanitation, food protection, and laboratory groups. Other important help was supplied by various district and local public health agencies, as well as the Sullivan County Public Health nursing services. Last, an epidemiology intelligence officer from the CDC in Atlanta rounded out the team. There was plenty of work for everyone. Says Kondracki, “We needed to establish the usual things: time, place, and person.”

  And they needed to establish these parameters quickly. The first round of phone calls showed that the scope of the problem was growing rapidly, with cases popping up throughout the state. Kondracki and his colleagues’ first steps were to design and distribute surveys to a large group of people. There were almost ten thousand firefighters who had stayed at six different hotels in the area, in addition to all of the other guests. Performing the survey and analyzing the data was no small task. The surveys quickly yielded an important piece of information. All of the typhoid patients had stayed in a single hotel (which I’ll call Grovers). “The hotel management was extremely cooperative,” recalls Kondracki. “We obtained a guest list for the period just prior to and immediately following the June 11–14 convention time.”

  Another salvo of telephone calls helped to narrow down the time frame. Only guests staying at Grovers during the convention developed typhoid, but not all of them. Kondracki had the time and place, but he still needed to pin down the exact persons. Why did some guests staying at Grovers between June 11 and 14 fall ill while others did not? To answer that question, the epidemiologists examined the two most likely potential sources—the water and the food. Grovers used the municipal water supply, which was quickly tested and ruled out as a source. The plumbing was also checked to make sure there was no cross-contamination between a sewer pipe and a clean water source. This also proved to be a dead end.

  The investigators had already surveyed the hotel guests about which meals they had taken at the hotel and which foods they had eaten at those meals. This was a formidable job: the hotel had 450 rooms with a capacity of 800 guests, as well as approximately 250 employees, half of whom work in food service. After analyzing the surveys, the investigators pinned down the particular meal that had caused the illness with surgical precision.

  All of the people who developed typhoid had eaten breakfast at Grovers on the morning of Tuesday, June 13.

  But there were still critical questions that needed answering. Roy Harvey had eaten breakfast at Grovers that morning, but so had Rita Harvey, and she was still perfectly well. After an initial rally, Roy’s condition deteriorated. By his third day in the hospital, his fever broke, and he began to feel better. He was even hoping to be discharged the following Monday, July 10. But over the weekend, he developed one of the feared complications of typhoid—intestinal bleeding. The inflammation had burrowed into the richly vascular lymphoid tissue in his small intestine. When the infection eroded deep enough into these blood vessels, he began to bleed. “When I looked in the toilet bowl, it was filled with bright red blood,” he remembers. “I called Heather, the nurse, and she helped me back to bed. I had lost about half my blood and needed a transfusion. I remember there were informed consent forms to sign.

  “I didn’t want to get the blood, but she said it was either that or possibly bleed to death. When she put it that way, the choice was easy. I got four pints,” said Harvey.

  The New York epidemiologists had their goal clearly in mind—to find the source and to prevent more cases. To accomplish this, the investigators still had two important tasks to complete. The first was to identify the specific item at breakfast that had been contaminated. The second was to find out just how the typhoid bacillus got there.

  Grovers’ food service was a large operation. There was a main kitchen and a coffee shop that, combined, served three thousand meals per day, with more than eighty menu items. The answer to the first question came rapidly; the food surveys distributed to the guests answered it unequivocally. But the food that the questionnaires implicated seemed very unlikely: orange juice. This was a surprise because the natural acidity of orange juice makes it a very unlikely vehicle to support the growth of S. typhi.

  The second job, to determine precisely how the typhoid got into the juice, remained elusive. Questionnaires would not solve this dilemma; that would require some old-fashioned shoe-leather detective work.

  Nearly all the hotel employees had cultures taken to test for typhoid. One worker, who was of interest to the investigation, had left Grovers shortly after the convention, on June 21. Of the remaining 250 workers, only one, a man I’ll call George Watkins, tested positive for S. typhi. Watkins was a dishwasher, but he occasionally helped prepare the orange juice, and he always drank several glasses each morning. Watkins’ stool cultures grew the bacterium only for about a week after the outbreak, and then they turned negative in follow-up tests, and remained that way. The transient positive result suggested that he may have been a victim of the outbreak, rather than its source. If he had been a carrier, he would most likely have remained culture positive, and his blood test for antibodies (which was negative) would most likely have been positive too. The worker who had left the employ of Grovers was a breakfast cook who had been in charge of making the orange juice. He did so by emptying twenty-four thirty-two-ounce cans of juice concentrate into a fifty-five-gallon plastic drum, similar to a large trash can. He would then add twenty-four gallons of water and stir the mixture with a wire whisk.

  Every typhoid outbreak begins with a human excreting the typhoid bacillus through the stool (or, rarely, in the urine). The infected stool might come from a person who is sick with or recovering from an acute infection with typhoid fever, or, more commonly, from a carrier of the typhoid bacteria. The carrier state is an interesting balance that is struck between pathogen (in this case, S. typhi) and host (in this case, someone who has been naturally infected with the typhoid bacillus). Roughly 3 percent of typhoid patients become carriers.

  The pioneering Prussian physician and microbiologist Robert Koch first suggested the concept of a typhoid carrier. He presented his hypothesis in a talk in Berlin on November 28, 1902, after observing that a small percentage of patients recovering from typhoid continued to shed the bacilli for months after recovery. The concept of a carrier was a novel, still unproved theory. Typhoid carriers persistently shed typhoid bacteria in their stools, even though they are not ill. If sewage contaminated with typhoid makes its way into drinking water, or if a carrier prepares food without thoroughly washing his hands after going to the bathroom, the bacillus can spread like wildfire.

  In the Aberdeen epidemic, some patients became sick, not from eating the contaminated corned beef, but from eating other meats that had been cut with the same knife!

  Without doubt the most famous typhoid carrier in history is “Typhoid Mary,” born Mary Mallon in Cookstown, County Tyrone, Ireland, on September 23, 1869. At the age of fourteen, the tall, stocky blonde moved to New York City, where she began working as a cook. Her employment record was abysmal; she hopped from job to job.

  Toward the end of August 1906, the daughter of a wealthy New York banker, Charles Henry Warren, developed typhoid fever. The family summered in fashionable Oyster Bay, Long Island, where they rented
a large property. When five other members of the household quickly fell ill with typhoid too, local health officials investigated. They found no contaminated food, water, or milk, so, having come up empty-handed, they closed the investigation.

  The owner of the house, concerned that he would be unable to rent the place the next summer, hired Dr. George Soper, a sanitary engineer with the New York City Health Department and a recognized typhoid expert, to investigate further. Soper followed standard procedures, first verifying the results of the previous investigation. There being no demonstrable source of the bacteria in any food or drink at the house, or in the plumbing (both indoor and outdoor), Soper considered the possibility of a typhoid carrier. He hoped to confirm Koch’s theory. He learned that the Warren family had recently hired a cook, one Mary Mallon, and he wanted to find her.

  One senses Soper’s excitement from his own description of that quest.

  First I went to the employment agency where I was given the missing cook’s former places of employment and the different people who had furnished her with references. Working from agency to agency I came in possession of little fragments of her history for ten years. What do you suppose I found out? That in every household in which she had worked in the last ten years, there had been an outbreak of typhoid fever. Mind you, there wasn’t a single exception.

  The question that confronted me now was: “Where is she?” Following her trail backward to cases in 1904, I found she had worked at the home of Henry Gilsey at Sands Point, Long Island, where four of seven servants suddenly got the disease. Going back still further, I found that five weeks after Mary had gone to cook at the home of J. Coleman Drayton at Dark Harbor, Maine, in 1902, seven out of nine persons in the house contracted typhoid, and so did a trained nurse and a woman who came to the house to work by the day. There had been an outbreak of the disease in New York in 1901, and I had reason to believe that Mary was behind this. In 1904, Tuxedo Park, the fashionable summer resort, was stricken, . . . and [I] discovered she had cooked there in that time.